Abstract The effects of non-standard visibility observation conditions upon computed visual range (meteorological range) were investigated using data from an eastern U.S. visibility study sponsored by the Electric Power Research Institute. Data collected by multi-wavelength telephotometry was specially marked in the field when non-standard conditions were observed. The non-standard conditions explored were those which violate the assumptions implicit in routine visual range calculations. These conditions include bright and dark clouds behind the target, bright haze, shadowed targets, and snow-covered targets. The errors caused by non-standard conditions were evaluated by comparing derived visual ranges and observed luminances along affected and non-affected viewing paths. The factors introducing error into visual range calculations during bright haze and bright cloud conditions were found to depend strongly on the relative degrees of light scattering in front of and behind the target. Large errors were noted in visual ranges derived for snow-covered targets if an intrinsic contrast of − 1.0 was assumed; a method for approximating intrinsic sky/target contrast was applied to correct these errors. Recommendations are provided for recognition of non-standard conditions and field flagging of data.